Journal of Environmental Quality 30:1154-1162 (2001)
© 2001 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
TECHNICAL REPORT
Ecosystem Restoration
Revegetation of an Abandoned Uranium Millsite on the Colorado Plateau, Arizona
E.P. Glennb,
W.J. Waugh*,a,
D. Mooreb,
C. McKeonb and
S.G. Nelsonb
a Environmental Sciences Laboratory, U.S. Department of Energy, 2597 B 3/4 Rd., Grand Junction, CO 81503
b Environmental Research Laboratory, University of Arizona, 2601 E. Airport Drive, Tucson, AZ 85721
* Corresponding author (jody.waugh{at}doegjpo.com)
Received for publication June 19, 2000.
 |
ABSTRACT
|
|---|
We attempted to restore native plants on disturbed sites at a former uranium mill on the Colorado Plateau near Tuba City, AZ. Four-wing saltbush [Atriplex canescens (Pursh) Nutt.] was successfully established in compacted caliche soil and in unconsolidated dune soil when transplants were irrigated through the first summer with 20 L/plant/wk. The caliche soil was ripped before planting to improve water-holding capacity. The diploid saltbush variety, angustifolia, had higher survival and growth than the common tetraploid variety, occidentalis, especially on dune soil. The angustifolia variety grew to 0.3 to 0.4 m3 per plant over 3 yr even though irrigation was provided only during the establishment year. By contrast, direct seeding of a variety of native forbs, grasses, and shrubs yielded poor results, despite supplemental irrigation throughout the first summer. In this arid environment (precipitation = 100 to 200 mm/yr), the most effective revegetation strategy is to establish keystone native shrubs, such as four-wing saltbush, using transplants and irrigation during the establishment year, rather than attempting to establish a diverse plant community all at once.
|
INTRODUCTION
|
|---|
ARID landscapes around the world have been adversely altered by mining that disrupts or eliminates native plant communities (Middleton and Thomas, 1997). Regulatory and land management agencies often require the reclamation of these sites following cessation of mining or milling activities. Reclamation is a return of the soil and plant community it supports to a condition of stability and productivity comparable with that prior to disturbance (Allen, 1988). Numerous methods to revegetate these sites have been attempted (reviewed in Bainbridge et al., 1993; Bleak et al., 1965; Call and Roundy, 1991; Cox et al., 1982; Day and Ludeke, 1986; DePuit, 1988; Jackson et al., 1991; May, 1975; National Academy of Sciences, 1974; Powell et al., 1990; Schaller and Sutton, 1978). Low-cost methods generally consist of seed and mulch applied to prepared soils without supplemental irrigation. Higher-cost methods include supplemental irrigation to promote seed germination and plant establishment (DePuit et al., 1982). The most expensive methods use transplants from nursery stock and intensive irrigation to produce a plant stand (Glenn et al., 1998; Grantz et al., 1998).
The choice of methods is difficult, because success or failure depends greatly on local edaphic and meteorological conditions. The underlying ecological processes that control vegetation composition at a specific site are usually unknown; hence, the best guidance for designing and implementing a revegetation program often is anecdotal evidence (Call and Roundy, 1991; Allen, 1995). The success of dry seeding methods generally declines as aridity increases (Call and Roundy, 1991). Some researchers consider irrigation essential in areas that receive less than 250 mm of annual precipitation (Aldon, 1978; Day and Ludeke, 1986; National Academy of Sciences, 1974), but the need for irrigation and the amount and application mode have been debated (DePuit, 1988).
We are attempting to establish native vegetation on abandoned uranium millsites on Navajo Nation land on the Colorado Plateau in the southwestern United States (Glenn et al., 1998; Lash et al., 1999). The present study was conducted at the Tuba City site in northeastern Arizona. Tuba City is one of many abandoned uranium millsites for which the U.S. Department of Energy was assigned responsibility for cleanup under the Uranium Mill Tailings Radiation Control Act (UMTRCA) of 1978 (U.S. Department of Energy, 1989a, b). Ten years before the present study, the Department of Energy replaced the topsoil around the site with sand from a nearby site, mulched and seeded the site with a selection of native plants, and then fenced the area to exclude grazing (U.S. Department of Energy, 1989c). Contaminated topsoil was buried on site with uranium mill tailings in an engineered disposal cell. Ten years after soil removal, replacement, and reseeding, the treated area had considerably less plant diversity than untreated areas (grazed and ungrazed) adjacent to the site (Lash et al., 1999).
We conducted a series of experiments for the Department of Energy using native shrubs, grasses, forbs, and two soils types, combined with standard revegetation methods (direct seeding, seeding plus irrigation, transplanting, and transplanting plus irrigation), to determine the best techniques for the Tuba City site. The native shrub four-wing saltbush can be established successfully using transplants and intensive irrigation during the establishment year. Direct seeding methods, direct seeding plus low-intensity irrigation, and transplanting plus low-intensity irrigation did not produce acceptable stands of any of the plant species used. The implications of these findings for other desert sites are discussed.
|
MATERIALS AND METHODS
|
|---|
Site Description
This former uranium millsite is located on the Navajo Nation, 8 km east of Tuba City, Arizona, just south of Highway 160 at an elevation of 1554 m (36°08'N, 111°10'W) (Fig. 1). The surrounding landscape is coppice dune, hummock, and swale desert topography overlying sandstone terraces that slope 0 to 8% to the southwest. The well-drained surface soils have moderately fine sandy textures and are part of the Badland–Torriorthents–Torrifulevents association (Hendricks, 1988). Dominant shrubs in this region are four-wing saltbush, shadscale [A. confertifolia (Torr. and Frem.) S. Watson], Mormon tea (Ephedra viridis Coville), banana yucca (Yucca baccata Torr.), and snakeweed [Gutierrezia sarothrae (Pursh) Britton & Rusby]. Two dominant non-native annual forbs are Russian thistle (Salsola kali L.) and bur ragweed (Ambrosia confertiflora DC.). Perennial grasses include Indian ricegrass [Achnatherum hymenoides (Roem. & Schult.)], sand dropseed [Sporobolus cryptandrus (Torr.) A. Gray], black grama [Bouteloua eriopoda (Torr.) Torr.], crested wheatgrass [Agropyron desertorum (Fisch. ex Link) Schult.] (a non-native), galleta grass [Hilaria jamesii (Torr.) Benth.], and muhly grass (Muhlenbergia pungens Thurb.). Annual grasses are six-weeks fescue [Vulpia octoflora (Walter) Rydb.] and foxtail brome (Bromus rubens L.) (a non-native). The range is heavily grazed by sheep and cattle (J. Willeto, Navajo Department of Agriculture, Window Rock, AZ, personal communication, 1997).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 1. Location of Experiment 1 Caliche Site and Dune Site plots and Experiment 2 Caliche Site plots at the Tuba City Uranium Mill Tailings Radiation Control Act (UMTRCA) Site, Arizona.
|
|
The initial remediation effort took place between 1986 and 1988 (Jacobs Engineering Group, Inc., 1995) and included fencing to exclude livestock. Inside the fenced area, the surface 25 cm of soil was scraped from approximately 10 ha of the former millsite to remove surface contamination that accumulated during milling operations. The contaminated soil and other radioactive materials were buried on site in a disposal cell beneath an engineered cover consisting of a compacted clay layer topped with gravel and rock (U.S. Department of Energy, 1989b). The topsoil in the scraped area was replaced with sand from a local source. The replacement material has chemical and textural characteristics similar to the surrounding soils (Lash et al., 1999).
Overview of Experiments
The revegetation experiments were located near the toe of the disposal cell (Fig. 1). Two soil conditions are present on the site. The first type is an undisturbed, coppice dune soil that is a fine to medium sand. The second type is a caliche soil that is present under the thin layer of surface sand that replaced the contaminated soil. The caliche soil is primarily sand and gravel cemented together by calcium carbonate precipitates and is nearly impenetrable in its natural state.
Experiment 1 attempted to establish two local varieties of four-wing saltbush, occidentalis and angustifolia, on the two soil types. The former, a tetraploid plant, is the most common variety of this species. The diploid var. angustifolia is a faster-growing form that appears to be specialized for rapid establishment in dune environments (Glenn et al., 1998; Sanderson and Stutz, 1994). Accessions of each variety were established as transplants from nursery stock in planting wells and were individually irrigated once a week with 20 L of water during the first summer only. The experiment was initiated in spring 1996; Glenn et al. (1998) report the preliminary results after the first year of growth. Here we report results after 3 yr of growth.
Experiment 2 attempted to establish a suite of plants on only the caliche soil through direct seeding and transplanting, using supplemental sprinkler irrigation during the first summer after planting. The plant list included six perennial shrubs, four grasses, and one forb (Table 1). This experiment was initiated in fall 1997 and evaluated in summers 1998 and 1999. Both experiments included a treatment in which the caliche soil was ripped before planting to open the soil to increase its water-holding capacity and to allow the roots to penetrate deep into the soil.
View this table:
[in this window]
[in a new window]
|
Table 1. Plant species and seeding rates used in Experiment 2. Seed rate is based on grams of pure, live seed as determined by germination tests prior to sowing.
|
|
Methods for Experiment 1
Glenn et al. (1998) present details of plant collection, propagation, planting, and irrigation. Fruits from 10 accessions (individual plants) of four-wing saltbush were collected on the site in 1995 and were used as source material for all four-wing saltbush plants used in these studies. Seven of the 10 accessions were subsequently identified as the common variety occidentalis, while three accessions were identified as var. angustifolia, based on leaf width, fruit size, and the presence or absence of specific flavanol compounds (Glenn et al., 1998; Sanderson and Stutz, 1994). Approximately 750 seedlings were transplanted into each of two locations in May 1996.
The Dune Site (Location 1) consisted of a series of low, sandy, coppice dunes that were within the site fenceline and had not been disturbed by either milling or subsequent remediation activities. The Caliche Site (Location 2) had been compacted and graded during milling and remediation activities and had a gravelly, sandy soil cemented by caliche (calcium carbonate) lenses. The Dune Site soil was classified as a loamy sand while the Caliche Site soil was a sandy loam; Glenn et al. (1998) present details of soil analyses. Four-wing saltbush transplants were spaced at 3-m intervals in nine adjacent blocks at each location. Each block contained 10 plants of each accession planted in random order.
At the Dune Site, plants were placed directly into unaltered soil. At the Caliche Site, soil was first ripped to 1.0 to 1.3 m depth with a bulldozer equipped with a ripping bar, and plants were transplanted within the rip lines. Ripping loosened the soil to facilitate planting and allowed penetration of irrigation water into the soil. A wire mesh cage was placed over transplants to protect them from rabbits and other wildlife. Each plant received a single dose of slow-release fertilizer and was irrigated once a week through the first growing season (May through September 1996) with 20 L of water. Protective cages were removed in 1997 as plants began to grow through the mesh. No additional water or fertilizer was applied.
Plants were randomly sampled in May 1997 to estimate growth and survival for the first growing season (Glenn et al., 1998). In July 1999, we determined the survival and canopy measurements of all plants. Canopy volume (V) was calculated from the north–south width (a), the east–west width (b), and the height (h) of each surviving plant, using the formula (Bonham, 1989):
 |
Methods for Experiment 2
This experiment was conducted to determine if a suite of plants from seeds and from transplants, including grasses, forbs, and shrubs, could be established in the caliche soil. A fall planting was used to allow plants to germinate or establish during the winter (period of maximum precipitation) before they were exposed to summer heat. The experiment was conducted in nine adjacent plots (20 by 27 m) on a disturbed, compacted area of the site near the Caliche Site location of Experiment 1 (Fig. 1). A randomized split plot experimental design was used (Fig. 2). Whole plots were used to compare three planting methods (transplants only, direct seeding only, and transplants and direct seeding combined). Three replicate plots of each planting method were prepared for a total of nine whole plots. Two split-plot treatments, one for soil preparation method, and the other for duration of irrigation, were included in each whole plot. The soil in each whole plot was ripped at 3-m intervals to produce nine ripped lines per plot; plants within the ripped lanes were considered to be part of the ripped treatment, while plants in the lanes between rip lines were in the unripped treatment. Following ripping, each plot was disked to create a seed bed. Each plot was subdivided into three sections containing three ripped lanes and three unripped lanes. Each subdivision received one of three irrigation treatments: (i) irrigation applied only one time (immediately after planting in fall 1996); (ii) irrigation applied immediately after planting, then weekly in May 1997 (four applications); (iii) irrigation applied immediately after planting, then weekly from May through September 1997 (18 applications). Irrigation was applied as a spray from a high-pressure hose (20 min per subplot per application). Flow rate was measured with an in-line flow meter and averaged 40.5 L/min (810 L per application, or approximately 0.45 cm of water applied to the soil). Each irrigation split plot contained three ripped and three unripped rows. Therefore, there were three replicates per split plot and six split plots that received direct seeding, for a total of 18 replicates per irrigation treatment.
View larger version (86K):
[in this window]
[in a new window]
|
Fig. 2. Split-split-plot experimental design used for Experiment 2 to test the interactive effects of planting method, irrigation, and ripping.
|
|
Seed was broadcasted onto the plots after disking in November 1996. Table 1 presents the plant species and seeding rates for the direct-seeded plots. A hand-operated, crank-type seed applicator was used for smaller seeds and hand-broadcasting was used for four-wing saltbush var. angustifolia fruits because they were too large for the mechanical seeder. The seeds were then bedded into the soil by dragging a section of chain-link fence material over the soil with a tractor.
Seedlings transplanted into experimental plots were: four-wing saltbush var. angustifolia and occidentalis, shadscale, greasewood [Sarcobatus vermiculatus (Hook.) Torr.], Mormon tea, and rabbitbrush [Chrysothamnus nauseosus (Pall. ex Pursh) Britton]. Four-wing saltbush varieties were established from seed collected on site and shadscale was started from cuttings taken from plants growing near the site; all other species were started from seed obtained from Maple Leaf Seed Co., Ephraim, UT. Maple Leaf did not specify the origin of seeds. Seeds or cuttings were started in a greenhouse in Tucson, AZ, in July and transplanted into plots in November 1996, following ripping, disking, and seed application. Transplants were placed in ripped rows only (no unripped treatment) at 3-m spacing, with the order of species determined randomly within each row. Each irrigation treatment splitplot contained three ripped rows and each row contained one of each of the five transplant species.
After the seeding and transplanting was completed, all plots received one irrigation, then were left unirrigated until May 1997 when irrigation treatments were resumed as previously described. Transplants were protected by wire cages during the first growing season. Planting success was evaluated in June 1998 (20 mo after planting) and in July 1999 (33 mo after planting).
For the 1998 evaluation, we used an occular point intercept method (Floyd and Anderson, 1982) to estimate plant cover within each ripped and unripped row in each plot. The point intercept device consists of two wooden frames on adjustable legs, one spaced 10 cm above the other. Each 0.5- by 1.0-m frame was strung at 10-cm intervals with monofilament lines forming two identical grids, each with 50 intersections. A sighting point was established by viewing down on intersections of the upper grid and alligning them with intersections in the same position on the lower grid. Percent cover was estimated as the proportion of sighting points (hits) intercepting a target (e.g., plant, soil, gravel). The point intercept frame was placed at one randomly chosen location centered in each ripped and unripped lane. This sampling protocol was determined after preliminary sampling of three sites per row in nine rows to estimate the coefficient of variation (84%) and the sample size needed to separate means at P < 0.05 with differences of approximately 20% in total plant cover (n = 162) (Sokal and Rohlf, 1997).
Final evaluation of the experiment was conducted in 1999, using a line-intercept method to estimate plant cover. This method was more rapid and encompassed more of the total experiment area than the point intercept method. A tape measure was tightly stretched the length of each row and the intercept of plant species or bare soil under the line was recorded. Transplants were evaluated separately in June 1999 for survival rate and canopy dimensions as in Experiment 1.
Statistical Analysis of Data
For Experiment 1, we compared the growth of two varieties of the four-winged saltbush, angustifolia and occidentalis, in response to soil type with a two-way analysis of variance. The logarithm of the individual plant volume was used as the dependent variable, with soil type (ripped caliche or dune) and plant variety (angustifolia or occidentalis) as fixed factors. Log transformation was used to normalize the distribution of data.
For Experiment 2, we used a three-way analysis of variance to examine the effects of seeding, irrigation, and soil preparation on the mean percent cover of the species that were seeded and on total plant cover. The surviving seeded species were four-wing saltbush var. occidentalis and angustifolia, blue grama [Bouteloua gracilis (Kunth) Lag. ex Griffiths], rabbitbrush, galleta grass, and Indian ricegrass. The sum of the percent cover of these species was used as the dependent variable. The three independent factors (treatments) were planting method (three levels: not seeded, seeded only, transplants plus seeding), irrigation (three levels: 0, 4, or 18 applications during summer 1996), and soil preparation (two levels: ripped and nonripped). Similar analysis was made of the percent cover of individual species where appropriate. To compare the volume of the transplants among species for the three plots containing transplants, we used a one-way analysis of variance on log-transformed data to normalize the distribution of the data.
Climate Records and Other Sources of Information
Temperature and precipitation data for Tuba City were obtained from the Western Regional Climate Center of the Desert Research Institute (wrcc@dri.edu) in Reno, Nevada. Plant identification was based on descriptions in Kearney and Pebbles (1960) and Benson and Darrow (1981), supplemented by a checklist of local species and on comparison with herbarium specimens for problematic spieces (Lash et al., 1999). Some plants could only be identified to genus level because flowers and fruits were not present. Globemallow (Sphaeralcea sp.) in the seed mix was also not identified to species by the seed supplier.
|
RESULTS
|
|---|
Meteorological Conditions during the Study Period
The site is characterized by hot summers (mean monthly temperatures greater than 20°C from June through August), cool winters (less than 4°C from December through February), and irregular precipitation (100-yr mean = 161 mm/yr) arriving mainly as winter and late summer rains. Spring is a period of drought. Figure 3 presents mean monthly temperature and precipitation data for the study period. Temperatures during the study were approximately normal; precipitation was much lower than the 100-yr mean in 1996 (56 mm) and higher than the mean in 1997 and 1998 (218 and 204 mm, respectively). Precipitation for January through August 1999 was slightly higher than the 100-yr mean value (125 mm compared with 102 mm). Experiment 1 was established during a period of relative drought, but the plants in that study were irrigated during the first summer. Experiment 2 was seeded during a relatively wet period.
Survival and Growth of Four-Wing Saltbush Varieties on Two Soil Types (Experiment 1)
For Experiment 1, transplants of four-wing saltbush were individually irrigated during the first summer. The ripped sandy loam soil at the Caliche Site most likely retained water around the root zone of the plants longer than the loamy sand at the Dune Site, in which water probably drained to deeper layers; but establishment and growth of both of the varieties were successful at both sites. Because the 1997 analysis (Glenn et al., 1998) showed no statistical differences in survival or growth of different accessions within each variety, data were pooled by variety for the 1999 evaluation. Varietal and soil-related differences first appeared in 1997 (Glenn et al., 1998) and remained evident in 1999 (Table 2); significant effects on the log-volumes of the plants were attributable to both soil type (F1,681 = 14.885, P = 0.000) and variety (F1,681 = 48.514, P = 0.000). Also, there was a significant interaction between these factors (F1,681 = 4.367, P = 0.037). Plants of both varieties grew best in the ripped caliche soil and those of var. angustifolia were larger than those of var. occidentalis, especially in the dune soil. Mean canopy volume across varieties was 0.3 m3 per plant at the Caliche Site compared with 0.2 m3 at the Dune Site; however, var. angustifolia had twice the canopy volume of var. occidentalis across sites. Both varieties had greater than 75% survival at the Caliche Site. At the Dune Site, var. angustifolia had 63% survival compared with only 33% survival for var. occidentalis.
View this table:
[in this window]
[in a new window]
|
Table 2. Comparison of the volumes of two varieties of four-wing saltbrush in two soil types in 1999, after 3 yr growth in Experiment 1.
|
|
Direct Seeding of Shrubs, Grasses, and Forbs (Experiment 2)
Treatments were first evaluated in June 1998, 18 mo after sowing. Plant cover was less than 15% in all treatments, and treatment effects were not significant for irrigation treatment or for planting method (seeded plots did not have higher plant cover than unseeded plots). Ripped lines had significantly (P = 0.003) greater plant cover than unripped lines (means are 10.7 and 7.6%, respectively), but for all treatments, differences in cover among species were all nonsignificant (P > 0.05). This was due to the very low cover of most species and the low power of the point-intercept measurement method, which was used to sample only a small portion of each row. We also considered that the seeds might require longer than 18 mo to germinate and establish sufficient plant cover to be measurable in the plots. Therefore, final evaluation was conducted after 30 mo using the line intercept method.
Twenty-seven species were encountered in the treatment plots in 1999 (Table 3), but overall plant cover was still very low when pooled for all treatments. Only 6 of the 12 species included in the seed mix were found in the plots. Treatments were initially evaluated based on total plant cover and only for those plants included in the seed mix. Total plant cover in all plots was dominated by alkalai sacaton [Sporobolus airoides (Torr.) Torr.], a bunch grass that was not included in the seed mix. The irrigation treatment was not significant (P > 0.05), but it was difficult to evaluate the other treatments because of the variability in the cover of alkalai sacaton among plots. Therefore, effects of planting and ripping were evaluated only for those plants included in the seed mix. The percent cover of the surviving seeded species differed significantly among the planting treatments (F2,117 = 3.311, P = 0.04) and among the soil preparation treatments (F1,117 = 6.259, P = 0.014), but there was no significant effect attributed to irrigation level (F2,11 = 1.313, P = 0.273).
Both seeding and ripping resulted in a higher percent cover of the seeded species, but the effect was not dramatic, as the percent cover was less than 6% in all cases (Table 4). Application of seed increased plant cover of the seeded species from 3.7 to 4.9%, while ripping increased the cover of those species from 2.7 to 4.3%. The top treatment (5.7% for seeded and ripped plots) was 3.4 times higher than the lowest treatment (1.7% for unseeded and unripped plots) for seeds in the seed mix. However, no treatment was effective in establishing a satisfactory plant cover on these plots.
View this table:
[in this window]
[in a new window]
|
Table 4. Mean and standard deviation (SD) of the sum of the percent cover of the six surviving seeded species (four-wing saltbush [var. occidentalis and angustifoia], blue grama, rabbitbrush, galleta, and Indian ricegrass). Percent cover was determined in 1999.
|
|
Individual species had a low percent cover and treatment effects were mainly nonsignificant (P > 0.05). However, alkalai sacaton was less abundant in ripped than in unripped rows (F1,117 = 64.181, P = 0.000), as expected, because ripping removed preexisting plants. Four-wing saltbush was the only seeded species that had significantly greater cover (F2,117 = 3.099, P = 0.049) in seeded plots.
Survival and Growth of Transplants (Experiment 2)
Survival and growth of all transplanted species in Experiment 2 were poor compared with those achieved with four-wing saltbush in Experiment 1. In Experiment 2, there were significant differences in the log-volumes of individual plants among species (F4,72 = 8.793, P = 0.000). Four-wing saltbush var. occidentalis were the largest plants followed by rabbitbrush (Table 5). The two varieties of four-wing saltbush had the best survival, both just more than 40%. Greasewood and shadscale plants exhibited both poor survival (3 and 10%, respectively) and poor growth.
|
DISCUSSION
|
|---|
The main conclusions of this study are that four-wing saltbush alone can be established in either dune or ripped caliche soil at this site when transplanted and irrigated individually, but that direct-seeding a mixture of native grasses, forbs, and shrubs is not likely to be successful with low amounts of supplemental irrigation. Our experiments typified two different approaches to revegetation: Experiment 1 concentrated on establishing a keystone shrub species at widely spaced intervals, while Experiment 2 attempted to establish a complete ground cover of shrubs, grasses, and forbs. Only the first approach was successful.
Call and Roundy (1991) point out that many land managers conceptualize revegetation as an instantaneous process where plants are expected to establish rapidly and form a permanent, static ecosystem. In reality, succession on arid rangelands is a slow, stochastic process and projects that use the approach taken in Experiment 2 often fail (Call and Roundy, 1991; DePuit, 1988; Lash et al., 1999). Numerous authors (e.g., Call and Roundy, 1991; DePuit, 1988; Grantz et al., 1998; Powell et al., 1990) have concluded that establishment of shrubs, as attempted in Experiment 1, is the most important first step in revegetation. Newly established shrubs provide shade and accumulation of litter creating microhabitats, or safe sites (Harper et al., 1965), that favor the germination and establishment of understory species. Our results support their conclusion.
Four-wing saltbush var. angustifolia transplants that were planted in ripped rows in caliche soil and individually irrigated through the first summer gave the best results among the shrub treatments (Experiment 1). These shrubs had 90% survival and achieved a ground cover of 0.4 m3 per plant after 3 yr. They increased sixfold in plant volume between 1997 and 1999, even though they received no irrigation after the first growing season. This increase in plant volume suggests that they established a root system capable of extracting soil moisture from greater depths in the vadose zone at this site (Jacobs Engineering Group, Inc., 1995). Variety occidentalis had 76% survival and achieved a ground cover of 0.2 m3 with the same treatment. Both varieties had lower growth and survival in dune soil than in caliche soil, because the dune sand probably has a lower water-holding capacity than the loam soil. However, there was a significant varietal interaction, with var. occidentalis having much lower survival in dune soil than var. angustifolia. This study supports the hypothesis (Sanderson and Stutz, 1994; Glenn et al., 1998) that the diploid, var. angustifolia, is specialized for rapid establishment and growth on shifting dune soils. It should be the preferred variety for revegetation on sites with a coppice dune, hummock and swale topography.
By contrast, in Experiment 2, plant establishment was largely unsuccessful. Direct seeding of shrubs, grasses, and forbs produced few established plants after 2 yr; percent cover of the seeded plants was <6%. In contrast, total plant cover in surrounding undisturbed areas was greater than 24%, and in other reseeded areas was greater than 19% (Lash et al., 1999). The success of direct seeding without irrigation is constrained by the low and erratic precipitation patterns of arid sites such as this one on the Colorado Plateau, where seedling recruitment may be successful only once every 15 yr (Bleak et al., 1965). Hence, we used supplemental irrigation to attempt to stimulate germination and establishment during the first summer. Our highest irrigation treatment supplied 7.2 cm of water over the first summer after seeding, effectively doubling the natural summer precipitation rate (6.9 cm). An examination of historical precipitation records shows that only one year (1914) of the past 100 for which there are weather records provided as much water in the May through August period as the 14.1 cm provided in the highest irrigation treatment (irrigation plus precipitation). Nevertheless, irrigation did not increase plant cover measured 2 yr later. Potential evaporation is as high as 1 cm/d during summer, and the plots dried rapidly between irrigations.
Other studies have reported successful seed germination in arid zones with supplemental irrigation (Ries et al., 1988; DePuit et al., 1982). Those studies, however, used irrigation rates ranging from 40 to 80 cm/yr over 1 or 2 yr. That amount of water was not available at our study site and seldom is available at desert revegetation sites. Using lower volumes of irrigation (as low as 5 cm per month during the first season), Powell et al. (1990) reported an initial stimulation of seed germination but no beneficial effects of irrigation that persisted over time in a cold desert climate. Ries et al. (1988) suggested that even small amounts of supplemental irrigation, which boosts total precipitation to amounts available in wet years at a given location, might be beneficial, but our results show that even a doubling of natural precipitation may not increase plant establishment in an arid climate.
Results with transplants in Experiment 2 were also disappointing compared with Experiment 1. Four-wing saltbush varieties had only 40% survival, even though they were in ripped soil of the same type as plants in the Caliche Site in Experiment 1. Rabbitbrush, greasewood, and shadscale all had survival rates of less than 50%, and all transplants had slower growth rates than plants in Experiment 1. The primary factors influencing the differences between experiments appear to be the amount of water applied and the irrigation method. Other possible factors include planting season and ambient precipitation amount. Transplants in Experiment 1 received 20 L/wk of water through the first summer, whereas (assuming each transplant harvested water from 0.25 m2 of plot area) transplants in Experiment 2 recieved only 2 L/wk of water. Grantz et al. (1998) reported similar poor results for shrubs irrigated with 2 L/wk of supplemental irrigation in the western Mohave Desert. Four-wing saltbush was the most successful species in their study, but they concluded that transplanting is effective only with large amounts of supplemental irrigation during the first summer.
Experiment 1 used less water per unit of land area than Experiment 2: 2.2 L/m2 compared with 4.5 L/m2. Water was applied individually to widely spaced shrubs in Experiment 1 but was broadcast over the entire plot in Experiment 2. Where the water supply is limited, the available water should be directed at individual shrubs rather than broadcast over the entire area requiring revegetation. Drip irrigation lines can be installed to distribute water to shrubs. In very arid climates like Tuba City, we recommend that the initial revegetation step should be to establish keystone shrubs. Four-wing saltbush var. angustifolia appears to be the most suitable selection for arid sites on the Colorado Plateau.
Transplant and irrigation methods are expensive compared with direct seeding because they require greenhouse propagation of plants, ripping of the soil, hand planting, and either installation of an irrigation system or hand watering. However, these methods produce better results and constitute a small amount of the total cost of remedial action at Uranium Mill Tailings Radiation Control Act (UMTRCA) Project sites where reestablishment of plant cover is considered essential for long-term success. Ten years after reseeding at the Tuba City UMTRCA Project site, shrub cover was sparse and total plant cover on seeded disturbed land was only 14% compared with 24% on undisturbed land (Lash et al., 1999). Our results support the conclusion of Call and Roundy (1991) that rapid establishment of a complete plant community on a disturbed desert site is difficult, but establishing shrubs that can facilitate the successional process is feasible. Whether the shrub community established in Experiment 1 will eventually lead to a diverse plant community of native species remains to be determined.
|
ACKNOWLEDGMENTS
|
|---|
This study was funded by the Long-Term Surveillance and Maintenance Project, U.S. Department of Energy Grand Junction Office, under DOE Contract Number DE-AC13-96GJ87335.
|
REFERENCES
|
|---|
- Aldon, E. 1978. Reclamation of coal-mined lands in the southwest. J. Soil Water Conserv. 33:75–79.
- Allen, E.B. 1988. The reconstruction of disturbed arid lands: An ecological approach. AAAS Selected Symp. 109. Am. Assoc. for the Adv. of Sci., Washington, DC.
- Allen, E.B. 1995. Restoration ecology: Limits and possibilities in arid and semiarid lands. p. 5–13. In Proc.: Wildland Shrub and Arid Land Restoration Symp., Las Vegas, NV. 19–21 Oct. 1993. USDA For. Serv. Gen. Tech. Rep. INT-315. Intermountain Res. Stn., Ogden, UT.
- Bainbridge, D., N. Sorenson, and R. Virginia. 1993. Revegetating desert plant communities. In T. Landis (ed.) Proc., Western Forest Nursery Assoc., Fallen Leaf Lake, CA. 14–18 Sept. 1992. USDA For. Serv. Gen. Tech. Rep. RM-221. Rocky Mountain For. and Range Exp. Stn., Fort Collins, CO.
- Benson, L., and R. Darrow. 1981. Trees and shrubs of the southwestern deserts. Univ. of Arizona Press, Tucson, AZ.
- Bleak, A., C. Frischnecht, A. Plummer, and R. Eckert. 1965. Problems in artificial and natural revegetation of the arid shadscale vegetation zone of Utah and Nevada. J. Range Manage. 18:49–65.
- Bonham, C. 1989. Measurements for terrestrial vegetation. John Wiley & Sons, New York.
- Call, C., and B. Roundy. 1991. Perspectives and processes in revegetation of arid and semiarid rangelands. J. Range Manage. 44:543–549.
- Cox, J., H. Morton, T. Johnson, G. Jordan, S. Martin, and L. Fierro. 1982. Vegetation restoration in the Chihuahuan and Sonoran Deserts of North America. Agric. Rev. and Manuals no. 28. USDA Agric. Res. Serv., Tucson, AZ.
- Day, A., and K. Ludeke. 1986. Reclamation and fertilization of coal mine soils in the southwestern desert. Desert Plants 8:20–22.
- DePuit, E. 1988. Productivity of reclaimed land–rangeland. p. 93–129. In L. Hossner (ed.) Reclamation of surface mined lands. Vol. II. CRC Press, Boca Raton, FL.
- DePuit, E., C. Skilbred, and J. Coenenberg. 1982. Effect of two years of irrigation on revegetation of coal surface-mined land in southeastern Montana reclamined for rangeland use. J. Range Manage. 35:67–74.
- Floyd, D., and J. Anderson. 1982. A new point intercept frame for estimating cover of vegetation. Vegetatio 50:185–186.
- Glenn, E., D. Moore, S. Sanderson, J. Brown, D. Lash, M. Nelson, and J. Waugh. 1998. Comparison of growth and morphology of Atriplex canescens varieties occidentalis and angustifolia. Southwestern Nat. 43:176–182.
- Grantz, D., D. Vaughn, R. Farber, B. Kim, L. Ashbaugh, T. VanCuren, R. Campbell, D. Bainbridge, and T. Zink. 1998. Transplanting native plants to revegetate abandonded farmland in the western Mohave Desert. J. Environ. Qual. 27:960–967.[Abstract/Free Full Text]
- Harper, J., J. Williams, and G. Sager. 1965. The behavior of seeds in soil I. The heterogeneity of soil surfaces and its role in determining the establishment of plants from seed. J. Ecol. 53:273–286.
- Hendricks, D. 1988. Arizona soils. Univ. of Arizona Press, Tucson, AZ.
- Jackson, L., J. McAuliffe, and B. Roundy. 1991. Desert restoration. Restoration Manage. Notes 9:71–80.
- Jacobs Engineering Group, Inc. 1995. Site observation and work plan for the Uranium Mill Tailings Remedial Action Program site at Tuba City, Arizona. U.S. Dep. of Energy, Albuquerque, NM.
- Kearney, T., and R. Pebbles. 1960. Arizona flora. Univ. of California Press, Berkeley, CA.
- Lash, D., E. Glenn, J. Waugh, and D. Baumgartner. 1999. Effects of grazing exclusion and reseeding on a former uranium millsite in the Great Basin Desert, Arizona. Arid Soil Res. Rehab. 13:253–264.
- Lindstrom, T.E. 1996. Forage and conservation planting guide. Utah State Univ., Logan, UT.
- May, M. 1975. Moisture relationships and treatments in revegetating strip mines in the arid west. J. Range Manage. 28:334–335.
- Middleton, N., and D. Thomas. 1997. World atlas of desertification. Arnold Press, London.
- National Academy of Sciences. 1974. Rehabilitation potential of western coal lands. Ballinger Publ. Co., Cambridge, MA.
- Powell, K., R. Vincent, E. DePuit, J. Smith, and F. Parady. 1990. Role of irrigation and fertilizer in revegetation of cold desert mined lands. J. Range Manage. 43:449–455.
- Ries, R., F. Sandoval, and J. Power. 1988. Irrigation water for vegetation establishment. J. Range Manage. 41:210–215.
- Sanderson, S., and H. Stutz. 1994. High chromosome numbers in Mohavian and Sonoran desert Atriplex canescens (Chenopodiaceae). Am. J. Bot. 81:1045–1053.
- Schaller, F., and P. Sutton (ed.) 1978. Reclamation of drastically disturbed lands. ASA, Madison, WI.
- Sokal, R., and F. Rohlf. 1997. Biometry. W.H. Freeman and Company, New York.
- U.S. Department of Energy. 1989a. Remedial action plan and site design for the stabilization of the inactive uranium mill tailings at Tuba City, Arizona. Final Rep. Uranium Mill Tailings Remedial Action Program, Albuquerque, NM.
- U.S. Department of Energy. 1989b. Technical approach document, Revision II. UMTRA-DOE/AL 050425.0002. Uranium Mill Tailings Remedial Action Project, Albuquerque, NM.
- U.S. Department of Energy. 1989c. Section 02935: Seeding, Tuba City Phase II. Document no. 4005a TUB-S-01-0055504. Uranium Mill Tailings Remedial Action Program, Albuquerque, NM.
This article has been cited by other articles:
|
|
|
|
 
S. Lutts, I. Lefevre, C. Delperee, S. Kivits, C. Dechamps, A. Robledo, and E. Correal
Heavy Metal Accumulation by the Halophyte Species Mediterranean Saltbush
J. Environ. Qual.,
July 1, 2004;
33(4):
1271 - 1279.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
